Author:
Balabanov P. V.,Divin A. G.,Egorov A. S.,Zhirkova A. A.
Abstract
The system of optical-electronic quality control of apples is described. An algorithm for detecting apple defects is proposed. It provides for obtaining information from a linear photodetector of a hyperspectral camera about the intensity of reflected light in the range of 400...1000 nm in 2.5 nm increments and subsequent processing of the obtained spectra, including the calculation of five vegetation indices. They are used as input parameters of a neural network designed to classify apple plant tissues by types of defects. The results of testing the system showed an accuracy of detecting defects of at least 87 %.
Publisher
Izdatel'skii dom Spektr, LLC
Cited by
1 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献